Traditional accounting-based frameworks

How does valuation in the Ohlson or Feltham-Ohlson framework differ from valuation in more traditional accounting-based frameworks? The traditional accounting-based valuation methods predict the future from information in the current financial statements.

These simple forecasts are based on the prediction that the current profitability and growth, as revealed in the financial statements, will continue in the future. In an SF1 forecast (where SF stands for simple forecast), earnings of the next period are forecast as the closing book value for the current period multiplied by the cost of capital.

Operating income is forecast by expecting the net operating assets to earn at the required return for operations. Finally, net financial expense is forecast by expecting the net financial obligations to incur the expense at the cost of net debt. As shown in table 1, SF1 assumes abnormal earnings to be nil. SF1's are good forecasts if the relevant balance sheet amount is at fair value. So an SF1 is typically a good forecast for the financing activities, but a poor forecast for the operating activities.

Since it predicts nil abnormal earnings for all future years, it is implied that the value of common equity is: Value of Common Equity = Book Value of Common Equity The SF valuations have the advantage of requiring little information and analysis of the future. They assume the future will be much like the present. All the information needed for an SF valuation is provided in the financial statements. They have poor explanatory power and lack of theoretical legitimacy: nothing supports the assumption that the future will look like the present.

SF valuations are therefore not very precise, but they serve as benchmarks, starting points to conduct more thorough analysis. Ohlson (1995) and Feltham-Ohlson (1995) attempt to build a more solid framework by means of specifying the relation between accounting data and firm value. However, in Ohlson's model, accounting is assumed to be asymptotically unbiased. This is not a realistic assumption. A more realistic model is provided by Feltham and Ohlson (1995).

Here the linear information dynamics is based on the following four linear recursive equations: In this framework, w12 captures the degree of accounting aggressiveness or conservatism. w12=0 unbiased accounting; w12>0 conservative accounting; w12<0 aggressive accounting. Additionally, w22 captures growth in net operating assets. Feltham-Ohlson shows that the intrinsic value of a stock also depends on the book value of net operating assets. Asymptotically, when growth and conservatism obtain, the PE ratio exceeds the normal ratio.